Anaplastic large cell neuroblastoma

Abstract

Anaplastic large cell neuroblastomas (ALCNB) are a subset of undifferentiated neuroblastomas with marked pleomorphic and anaplastic features that render them diagnostically challenging. We reviewed the records of all patients diagnosed with ALCNB at Children's Healthcare of Atlanta (Egleston Children's Hospital) for their clinical, biologic, and pathologic characteristics and their treatment outcomes. From 1998 to 2006, 7 patients were diagnosed with ALCNB. All patients presented with abdominal-pelvic masses, 3 of them of adrenal origin and 2 with thoracic extension, with clinical stages 3 or 4, and were considered to have high-risk disease. The N-MYC oncogene was amplified in 3 cases and catecholamines were elevated in 5 of 6 patients tested. All pretreatment tumors demonstrate pleomorphic, anaplastic morphology with bizarre mitoses admixed with undifferentiated but monomorphic cells with minimal if any neuropil or neuro-ganglionic differentiation. Immunohistochemical markers for neuron specific enolase (NSE) and synaptophysin were strongly positive in all specimens and chromogranin in 4 of 5. Interestingly, all tumors showed strong Fli-1 nuclear positivity despite a negative CD-99 stain. However, reverse transcription polymerase chain reaction or fluorescent in-situ hybridization testing for Ewing sarcoma transcripts was negative in 4 available specimens. This same Fli-1 antibody had tested negative on 30 conventional neuroblastomas, indicating a peculiar cross reactivity with this subset of ALCNB. Posttreatment biopsies showed maturation changes to more conventional neuroblastoma histology in 5 of the 7 cases. Follow-up ranged from 9 months to 4 years from diagnosis (median: 25 months). Five patients are still alive after treatment, 1 died 9 months after diagnosis, and another patient refused high-risk therapy and progressed and died 9 months from diagnosis. Anaplastic large cell neuroblastomas are a subset of undifferentiated neuroblastomas characterized by the absence or marked paucity of histologic clues for the diagnosis of neuroblastoma. Although all these tumors are strongly positive for NSE and synaptophysin, they also show Fli-1 positivity. However, they are negative by molecular testing for EWS transcripts, and they are immunohistochemically negative for CD99. The true neuroblastic nature of these tumors is supported by the N-MYC oncogene amplification in some of them, catecholamine production, immunohistological reactivity, and their posttherapy maturation to a more recognizable neuroblastic morphology. Although follow-up is still somewhat limited, the response and survival of the patients in our institution is better than a previous European study that indicated an aggressive clinical behavior of these tumors, although treatment modalities were not described in that report. Further study of this variant of neuroblastoma with more patients is required to determine optimal therapy, more accurately predict outcome, and to ascertain if ALCNB are a distinct biologic group of neuroblastomas.